The long-term goal of this work is to understand the biophysical basis of specific recognition and communication between signal transduction proteins. Malfunctions in molecular communication play a role in a wide variety of disorders, including cancer, inflammation, and endocrine disease. The interactions of many signaling proteins are mediated by modular protein-protein recognition domains. We have chosen to focus on one such module, the src-homology 3 (SH3) domain, in order to gain a detailed understanding of hQw it functions. These domains bind specific proline-rich sequences found in target proteins. Remarkably, initial structural characterization has revealed that the SH3 binding surface can recognize peptides in two possible, orientationally opposite modes. This conformational reversibility makes deciphering the recognition code of SH3 domains particularly challenging: in addition to identifying the sequences that bind, one must also identify their binding conformations. This proposal describes a strategy to elucidate the structural and energetic mechanism by which SH3 domains read and recognize specific amino acid sequences. The majority of this work will be performed on an SH3 domain from the C. elegans signaling protein, Sem-5. We shall use a combination of structural, biochemical, and molecular genetic approaches to test models of recognition, and to tease apart the energetics of this compact, but complex interacting system. The aims of this proposal are to: (l) Elucidate the role of packing complementarity in sequence and orientational discrimination; (2) Define the interactions that determine the sequence preferences of individual domains; (3) Utilize our understanding of SH3 interadtions to rationally design a novel class of peptidomimetic inhibitors; and (4) Elucidate how SH3: domains, can be used as building blocks to construct larger proteins that exhibit more complex binding and signaling properties.